基于回声状态网的光伏发电量预测

摘要

For the photovoltaic power generation system, analyzing the characteristics and factors of power generation, the photovoltaic power generation forecasting model is established based on echo state network. Using the historical data of the photovoltaic power generation system (such as electricity, light intensity, temperature) as learning samples, the model utilizes the simple and unique learning method of echo state network to learn network , and the stabile model is used to forecast the power generation. Simulation experiments show that the forecasting method can get higher prediction precision.%对光伏发电系统的发电特性及影响发电的因素进行分析,建立了基于回声状态网的光伏发电量预测模型. 该模型利用回声状态网的简单而独特的学习方法,将光伏发电系统的历史数据(如发电量、 光照强度、温度)作为模型的学习样本,对网络模型进行学习,并利用学习稳定的模型对发电量进行预测. 仿真实验表明,该预测方法能得到较高的预测精度.